Resolution Limits of Near-Field Electromagnetic Imaging

Low frequency electromagnetic tomography is a non-destructive imaging technique that utilizes low frequency (< 100 kHz) signals to image objects based on observed sensor data at a standoff distance. However, given the several kilometer wavelengths associated with low frequency signals, sensors are typically in the near-field region of the object being imaged. From a measurement point of view, near-field imaging inherently results in very similar measurements across sensors. This blurring effect determines the maximum possible resolution of a given near-field imaging system. In this work, we describe how to determine this resolution limit by replacing the object being imaged with an equivalent array of infinitesimal dipoles. Using this formalism, the scattered fields can be modeled by a matrix vector product; we then use a numerical rank analysis to determine the maximum resolution for which the dipole array can be inverted. Using this approach, we demonstrate how the resolution limit of near-field imaging systems can be determined based on sensor geometry, dyadic Green's functions, and other a-priori information.